CBT for difficult-to-treat depression: self-regulation model

Abstract Background: Cognitive behavioural therapy (CBT) is an effective treatment for depression but a significant minority of clients do not complete therapy, do not respond to it, or subsequently relapse. Non-responders, and those at risk of relapse, are more likely to have adverse childhood experiences, early-onset depression, co-morbidities, interpersonal problems and heightened risk. This is a heterogeneous group of clients who are currently difficult to treat. Aim: The aim was to develop a CBT model of depression that will be effective for difficult-to-treat clients who have not responded to standard CBT. Method: The method was to unify theory, evidence and clinical strategies within the field of CBT to develop an integrated CBT model. Single case methods were used to develop the treatment components. Results: A self-regulation model of depression has been developed. It proposes that depression is maintained by repeated interactions of self-identity disruption, impaired motivation, disengagement, rumination, intrusive memories and passive life goals. Depression is more difficult to treat when these processes become interlocked. Treatment based on the model builds self-regulation skills and restructures self-identity, rather than target negative beliefs. A bespoke therapy plan is formed out of ten treatment components, based on an individual case formulation. Conclusions: A self-regulation model of depression is proposed that integrates theory, evidence and practice within the field of CBT. It has been developed with difficult-to-treat cases as its primary purpose. A case example is described in a concurrent article (Barton et al., 2022) and further empirical tests are on-going.


Introduction
The goal of the self-regulation model is to provide effective therapy for depressed clients who have not responded to standard cognitive behavioural therapy (CBT) protocols, such as cognitive therapy (CT; Beck et al., 1979) and behavioural activation (BA; Martell et al., 2010). Nonresponders to CBT, and those who have heightened risk of relapse, form a diverse and heterogeneous group. They are often non-responsive to other psychological therapies, antidepressant medication and social interventions. They are sometimes described as clinically complex, treatment-resistant, or, more recently, difficult-to-treat has been proposed as a better way to describe their depression, at least with respect to current treatments (McAllister-Williams et al., 2020;Rush et al., 2022). Compared with treatment responders, they are more likely to have adverse childhood experiences, early-onset depression, co-morbidities, interpersonal problems and heightened risk to self (Wojnarowski et al., 2019). Recurrent episodes and chronic presentations are common, as are referrals to secondary and tertiary services.
Should clients who have not improved following an adequate dose of high-fidelity CBT be offered a different evidence-based treatment or a second course of CBT? This is a common dilemma across the stepped care system. Our response has been to adjust CBT to the needs of this client-group and we present the model we have developed in this article. The method we applied is collaborative and integrative, in two senses. Firstly, all interventions are guided by individualised case formulations in collaboration with clients, either in routine practice or single case research (Barton et al., 2008). When at all possible, treatment dilemmas are shared with clients and their feedback is used to shape model development and treatment strategies. Secondly, within the field of CBT we draw on several theories of depression, relevant cognitive science and treatment strategies, including first, second and third wave therapies. The resulting model is an integrated CBT approach. More information about the model development is available in Barton and Armstrong (2019).
While the self-regulation model can be used with any depressed client, therapy for difficult-totreat cases is its primary purpose. At present, it has a small, emergent evidence base and should not be used instead of CT or BA in the first instance; rather it can be considered when CT and/or BA have not had a sufficiently beneficial or lasting effect. Therapists use core CBT skills as they would with any CBT model: there are no special procedures that deviate from core competencies. It is a high dose treatment with up to 30 individual sessions, usually over a 12-month period. This is expensive compared with standard protocols but, when effective, it has the potential to save on future healthcare costs which are particularly high in this client-group (Johnston et al., 2019). There are ten treatment components and these are combined in a bespoke way, based on the needs of the individual undergoing therapy. Some of the components fall within established CBT practice, for example, alliance building, risk reduction and relapse prevention; some have overlapping features with second-and third-wave models such as mindfulnessbased cognitive therapy (MBCT; Teasdale et al., 2000) and rumination-focused cognitive behavioural therapy (RFCBT; Watkins et al., 2011); others are novel, for instance selforganisation, which seeks to influence networks of self-states rather than core beliefs. The model is grounded in self-regulation theory. It shares the same assumptions about selfregulation as originally described by Carver and Scheier (1999): (1) Personal identity is plural and based on multiple self-representations that provide selfdefinition across different situations.
(2) Specific 'selves' are goal-directed and motivated: they are seeking to approach or avoid different outcomes and respond to feedback in pursuit of their goals.
The model is not derived from all of Carver and Scheier's (1999) proposals; rather it shares the same fundamental assumptions about self-regulation. The model is a set of hypotheses about how self-regulation becomes dysregulated during depression and this is used to formulate individual cases and shape treatment.
This article introduces the model and treatment components and provides an example of a case formulation. It also explores similarities and differences compared with standard protocols, such as CT and BA. A concurrent article provides a case example of how the model is used in practice, and readers are encouraged to refer to that article to explore the therapeutic applications (Barton et al., 2022). The current evidence base is summarised in Barton and Armstrong (2019).

Self-identity disruption
A fundamental claim of the model is that depression is associated with disrupted or underdeveloped self-identity. Normal self-identity is not unitary or simple: it is constituted by multiple self-representations that develop at different points in the lifespan (Schwartz and Petrova, 2018). Some self-representations are internalised in early childhood, for example, the felt-sense of being loveable, competent and safe. These representations provide a structure within which life skills can develop, for example, interpersonal, social, educational, occupational, etc. Other selfrepresentations result from experiences in adolescence and adulthood when people become identified with particular goals, relationships, roles and group memberships (e.g. family, partner, social, relationships, work, occupation, interests, leisure, nationality, gender, spirituality, etc;Meca et al., 2019). As an individual develops, their self-identity usually becomes more plural and coherent, but for some people this is delayed or obstructed, leaving them with an unconsolidated or confused sense of self that can limit and/or complicate their development.
Self-representations vary in how core or peripheral they are and this can vary over time. When self-representations are core and self-defining, negative affect is likely when they are disrupted or destabilised. Disruptions can be precipitated contextually, for example, negative life events or developmental changes ('outside in'), or intra-psychically, for example, through thoughts, images or memories ('inside-out').
Vulnerability to depression is formulated as one or both of the following: (1) Limited internalisation of positive self-representations, particularly in childhood and adolescence.
(2) Narrow investment in specific positive self-representations, particularly in adulthood.
In this context, 'positive' means a self-representation that provides self-definition, value and purpose. In the cognitive model of Beck et al. (1979), vulnerability to depression derives from negative self-beliefs that result from adverse childhood experiences. In the self-regulation model, vulnerability to depression derives from the under-development of positive self-representations. This can be the result of early adverse experiences which can make it difficult for young people to develop positive self-identities (Elrefaay et al., 2021). When positive self-representations are under-developed, it is more difficult to successfully negotiate developmental tasks, such as adolescence. This is the period of life when most first-onset depressions occur, and people with early-onset depression are more prone to relapse and recurrence later in the lifespan (Bockting et al., 2015;Wojnarowski et al., 2019).
Depending on how adult self-identity is structured and organised, vulnerability to depression can increase in mid and later life, even when childhood experiences were positive and secure. Vulnerable individuals rely on a restricted range of positive self-representations and they are prone to over-investing in these, for example, an idealised relationship, career or over-valued goal (Lam et al., 1996). These can provide self-definition, value and purpose in the short-tomedium term (so long as they remain stable), but there is a hidden vulnerability to depression when self-identity has limited breadth and flexibility.

Depressed mood and impaired motivation
Depressed mood is a normal consequence of disrupted core self-representations. For most people, depressed moods are short-lived states that self-correct automatically or through the use of strategies. In major depression, there is a failure of self-correction: mood repair strategies (deliberate or automatic) are either not used, not sufficiently potent or are working against the intended outcome. In this situation, negative affect intensifies and positive affect weakens. The model describes a number of psychological processes that can interact repeatedly to create a downward spiral into depression, as depicted in Fig. 1.
When depressed mood persists, approach motivation is impaired: there is reduced positive anticipation, lower reward expectancies and weakened impulses towards desired states (Frey and McCabe, 2020;Kumar et al., 2018;Sherdell et al., 2012;Wu et al., 2017). Impaired approach is associated with reduced openness to new experiences and a defensive orientation when interacting with others (Treadway et al., 2012;Trew, 2011). Drive motivation, such as goal pursuit and reward seeking, is attenuated, and the motivation to self-soothe and relate to others can also be weakened (Watson et al., 1999). The net result is impoverished interest in goal-directed activity, which can be task-related, interpersonal or intra-personal. As approach impulses reduce in intensity, avoidant impulses tend to increase with heightened urges to avoid, escape or hide (Hershenberg et al., 2017;Roskes et al., 2014).

Rumination
Disengagement tends to increase self-focused attention and this is one of the setting conditions for rumination (Cribb et al., 2006;Koster et al., 2011;Lyubormirsky and Nolen-Hoeksema, 1995;Marroquín et al., 2010). The depressed mind has limited reflective capacity and it becomes fused and enmeshed with its negative contents (Watkins and Teasdale, 2001). This contrasts with reflective processing, which enables an individual to pay attention flexibly and de-centre from their mental events. The model proposes that depressed individuals are attempting to think clearly, understand events and make decisions, but the effects of self-focused attention and depressed mood compromise normal cognitive processing. Efforts at reflection and strategic thinking fall foul of the cognitive effects of depression, leading to brooding on repetitive themes and unproductive cycles of self-analysis (Nolen-Hoeksema, 1991). This is sometimes interspersed with phases of cognitive avoidance and emotional suppression, which at best provide temporary escape from the distress associated with rumination (Liverant et al., 2011).
The model proposes that rumination results from mismatches between goals and cognitive processes: the goal the person is trying to attain is mismatched to the process intended to achieve it (Watkins and Roberts, 2020). Concrete examples are observed in depressing questions, such as, 'why should I bother going on?'. The implicit goal in this question is to recover value and purpose, but this is defeated by the cognitive effects of the question which either lead to no answer or unhelpful answers.

Intrusive memories
The interaction between rumination and identity disruption results in positive memories becoming less accessible (Brewin, 2006) and negative memories becoming more accessible (Brewin et al., 1999;Gaddy and Ingram, 2014;Mihailova and Jobson, 2020). A significant proportion of these are intrusive, that is, unwanted recollections rather than the product of deliberate recall, and this is more pronounced in individuals with trauma histories (Mihailova and Jobson, 2018;Payne et al., 2019;Starr and Moulds, 2006). Intrusive memories are often subject to rumination: explanations of the past are sought but, as above, questions generated during depression tend to defeat their intended goals (e.g. 'Why do bad things always happen to me?'; Rosebrock et al., 2019;Wisco and Nolen-Hoeksema, 2009). In contrast to intrusions, deliberate recollections of past events are over-general and lack specificity. Over-general memory can be very frustrating for depressed individuals: in spite of cognitive effort, they struggle to access details to remember and learn from their past (Hallford et al., 2020). This is also associated with rumination and is pronounced in individuals who suffered childhood abuse (Griffith et al., 2016;Liu et al., 2017). This creates another vicious spiral where the excess of distressing intrusive memories fuels rumination, and the individual struggles to access the specific memories that would support reflective processing.

Passive life goals
The interaction between impaired motivation and identity disruption results in life goals becoming more passive. Rather than being tangible concrete possibilities, life goals become abstract wished-for states. Belief in being able to influence them is reduced and there is increased pessimism about achieving them (Dickson et al., 2011;Dickson et al., 2016). The net effect is increasing goal discrepancies: depressed individuals make minimal progress towards their life goals because they are not engaged with them, and this becomes an input for rumination and discrepancy-based thinking (e.g. 'How come I never get what I want?').
Because interactions with the environment are limited, there is minimal feedback on progress and it is difficult, when depressed, to assess whether goals are realistic. Consequently, depressed individuals are sometimes over-invested in goals, and possible selves, that are unlikely to be attainable (Lam et al., 1996). Life goals also tend to be on hold until other conditions are satisfied ('when I feel better'), which limits engagement with actions that could otherwise help to enhance mood (Coughlan et al., 2017;Hadley and MacLeod, 2010). Short-term goals shift to preventing depression from worsening rather than taking good risks to improve it. Even when life goals are active, and appear to be approach-based, there can be avoidant elements in the underlying motivation that subtly maintain vulnerability to depression (Sherratt and MacLeod, 2013).

Summary
The model proposes that depression is perpetuated by repeated interactions of self-identity disruption, impaired motivation, disengagement, rumination, intrusive memories and passive life goals. The key issue is the repeated interaction of these processes which captures the system like a traffic gridlock, in which the potential for change in any specific process is limited by the presence of other processes (Teasdale and Barnard, 1993). This has an entrapping effect and as depression becomes more severe, helplessness and hopelessness often develop as a consequence (Maier and Seligman, 2016;Siddaway et al., 2015). Figure 2 presents a case formulation of a 52-year-old married woman suffering with severe recurrent depression (pseudonym, Evelyn). Evelyn suffered adverse childhood experiences and these limited the development of positive self-representations, particularly in the interpersonal domain of being likeable and loveable. However, as a young person she experienced some degree of self-worth feeling responsible for others and performing well at school.

Case example
The self-regulation model focuses on positive self-representations: whether or not they were internalised across the lifespan, and how they influence current self-identity. Evelyn's selfidentity was narrowly invested in taking responsibility for others through her work as managing director of a company. This was a positive self-representation that provided selfdefinition, value and purpose. Evelyn was highly invested in the goal of doing her job well and valued this goal. This does not mean that she had consistently positive beliefs about doing a good job; in fact, these fluctuated a great deal. She experienced phases of work going reasonably well when she reported her mood to be 'OK', but small set-backs at work (real or perceived) had a disproportionate effect on her mood. There was nothing abnormal about being a managing director or the goal to help others at work; the problem was that Evelyn's self-identity was narrowly invested in the work domain and, as such, her global self-identity was prone to disruption when work-related setbacks occurred. Over-investment led to overgeneralisation. She could switch rapidly into a self-loathing, unmotivated, withdrawn, ruminative state, pre-occupied with memories of letting others down and uninterested in planning for the future. Evelyn's therapy is described in detail in Barton et al. (2022).

Treatment components
The aim of treatment is to enhance self-regulation skills and encourage the re-organisation of selfidentity. There are ten treatment components, as listed in Table 1.
An advantage of treatment protocols is standardising therapy to maximise consistency and therapist fidelity (McGlinchey and Dobson, 2003;Waltz et al., 1993). A disadvantage is providing the same interventions in the same sequence and dose for all clients, irrespective of client need. The model proposes that, for difficult-to-treat cases, there is an advantage in allowing individualisation of the components, provided that therapists maintain good fidelity to them. Individualisation means varying their sequence, combination and dose, based on client-need. It does not mean drifting from the therapist actions prescribed within them (Weck et al., 2011a;Weck et al., 2011b). This strategy is a way of responding to the heterogeneity observed in this client group. For example, some depressed clients have extensive trauma histories and are highly ruminative: their treatment needs to target intrusive memories, rumination and identity disruption. Other clients are stuck in the transition between adolescence and adulthood: their treatment needs to target impaired motivation, passive life goals and identity formation. Encouraging a bespoke plan is a way of balancing different client needs within the same treatment model.

Alliance building
In common with all CBT therapies, the therapist and the client need to form a collaborative working alliance (Cameron et al., 2018). In difficult-to-treat cases, the presence of multiple biomedical, psychological and social problems can be a barrier to this, creating confusion about what to target and sometimes leading to mistrust, excessive formulation or a lack of focality (Barton et al., 2017). Self-dysregulation can contribute to these difficulties, with clients struggling to engage, agree goals and contribute to shared reflections. It is important that alliance barriers are identified and made explicit so that they can be overcome, creating an optimal balance of support and change. The treatment dose of up to 30 sessions is intended to give clients and therapists the time they need to form and maintain a strong working alliance.

Treatment rationale
The treatment rationale is based on mood fluctuations. Some clients believe that their mood is invariant, but this is never the case: a sufficiently sensitive mood diary always reveals some variability, however slight. Therapists inquire about specific client experiences, some when their mood was very depressed, and others less so. The aim is to increase the client's awareness of the contrast between different mood-states. To learn from these, detailed information needs to be elicited about self-regulatory processes: affect, motivation, behavioural engagement, cognitive processes, memories, goals, etc. This helps to build the case formulation which is usually formed gradually over a number of sessions (e.g. Fig. 2). The emphasis on mood contrasts helps clients to recognise that they already experience less-depressed moods to some extent, even if they are shortlived and difficult to remember. To effect change, sufficient emphasis has to be placed on lessdepressed experiences, not just the maintenance formulation.

Approach motivation
The aim is to stimulate approach impulses and weaken avoidant impulses. Achieving this is centred on increasing attention on the client's needs and desires to make these explicit, forming intentions and making plans, and fostering active engagement. Questioning the client about what they felt like doing and were hoping would happen during less-depressed moods increases awareness of needs and desires. This is essential because everyday actions need to be approached in a deliberate intentional way at first, gradually becoming more automatic as normal self-regulation is restored. This applies to tasks (e.g. making a lunch), interactions with others (e.g. socialising with friends), and to interactions with oneself (e.g. self-soothing when distressed). Concretely, when considering situations prospectively, therapists ask: 'What do you need from this situation? What do you desire? What would you like to happen?'. This contrasts with standard CT, which usually asks: 'What do you expect or predict will happen?'. The explicitness of this process helps to generate reasons for action and, depending on the outcomes of active engagement, these gradually become motivational impulses. As motivation in daily life becomes stronger and more automatic, the client is encouraged to reflect on higher-order needs and deeper desires, to form an individualised hierarchy that guides their decisions and personal development.

Active engagement
The goal of active engagement is to increase clients' interaction, including daily tasks, relating to others and oneself. Engagement is targeted in areas where the client is currently disengaged, withdrawn or avoidant. There is curiosity about how clients interact, and there is a big emphasis on experimentation, with clients encouraged to try out new ways of relating and interacting, grounded in their needs and desires. When particular relationships or tasks are considered, the aim is to elicit concrete details about what the client would like to happen, aligned with their approach motives. All experiments are exploratory, to find out if the client can influence their preferred outcomes. Ways of actualising their preferences are explored in detail: how they will approach the situation, how they will interact, where they will place their attention, etc. At every point, it is essential to calibrate actions to what is manageable within the client's energy and window of tolerance. The aim is confirmatory learning: for clients to learn that they can influence their preferences sufficiently for it to be worth the effort and risk of doing so. They do not learn that life is fully controllable or predictable; rather, that seeking to influence preferred outcomes is a helpful way to self-regulate and approach different situations.

Mental freedom
The aim of mental freedom is to develop a good self-mind relationship with reflective capacity, attentional and questioning skills. The first step is to increase awareness of the difference between rumination and reflection. Positive beliefs about rumination need to be addressed, so it is recognised as an understandable but unhelpful process. In contrast, reflective processingwhich can address the same subjectively important contenthelps to generate new ideas, answer questions, reach conclusions and make decisions. Becoming aware of the difference between the two is itself a reflective activity (Arditte and Joormann, 2011). As reflective capacity increases, clients develop greater attentional skills, choosing where to place their attention in different situations. This is amenable to cognitive experiments in session, for example, varying the object and depth of attention to discover the effects on cognition and mood (e.g. internal vs external; mind vs body; Koster et al., 2017). Clients learn the value of external focus when interacting with tasks and other people, and this can be incorporated into active engagement. During self-reflection, they learn the value of paying attention to somatic and affective experiences, so that the body receives sufficient attention, not just the mind (Dey et al., 2018;Watkins and Moberly, 2009). As clients' cognitive skills increase, they learn to differentiate helpful and unhelpful questions through cognitive experiments: encouraging helpful questions is a way of counteracting the goal-process mismatches that are common in depressive rumination.

Self-organisation
The aim is to strengthen, diversify and re-structure positive self-representations. When this is effective, clients come to a fuller appreciation of their personal qualities and capacities (i.e. true self), and this acts as a buffer against the negative self-beliefs that often occur during depression (i.e. false self). Clients do not always recognise their personal qualities or internalise them as selfrepresentations. Self-organisation makes clients' qualities explicit, sometimes by involving family and friends in selected sessions, and encourages their consolidation and internalisation. When negative self-beliefs obstruct this, a theory A/B method is used to help the client de-centre from them. Theory A is the false self: e.g. 'I'm useless and worthless because that's how I feel'; theory B is an alternative explanation why A sometimes feels true: e.g. 'I sometimes feel useless and worthless because I was neglected and unfairly treated as a child. I internalised those experiences as if they were to do with me, rather than how I was treated. When I get depressed, it brings back those feelings'. Unlike standard CT, the content of negative selfbeliefs is not elaborated in detail: rather theory B is used to help the client de-centre from negative beliefs when they are present, so they are more able to engage with recognising and building positive self-representations.

Goal organisation
The aim is for clients to structure their life goals so they are attainable and linked to positive selfrepresentations. It is best for these to span a number of domains rather than narrow investment in one area (e.g. family, partner, social, relationships, work, occupation, interests, leisure, etc.). Mapping goals out explicitly can help to overcome specific problems, for example, when there are too few or too many goals, when they are avoidant rather than approach-based, when they are in conflict with each other, or highly conditional. Life goals need to be concrete, specific and imaginable, not abstract possibilities outside of the client's influence. Some clients also need structured support with planning, to break goals into sub-goals and concrete steps. This can help to reduce goal conflict and conditionality, and further encourage approach motivation and active engagement.

Memory integration
There are different aims for positive and negative memories. For negative memories, much depends on the personal significance of the event. If it is a minor, time-limited event, it is usually best to limit the attention paid to it, mitigating the risk of rumination elaborating the memory in an unhelpful way. However, it is important to differentiate memory avoidance from the freedom to place attention where it is needed: for example, 'I don't have to dwell on this' is mentally freer than 'I must not think about it'. For significant negative events, it is important that there is an opportunity to emotionally process the experience, to reduce the likelihood of it becoming unprocessed and intrusive in the future. This should be approached with a clear timeline, experientially and reflectively, grounded in concrete experiences and not in an abstract or analytic way (Hitchcock et al., 2017). The same approach is taken with intrusive memories that pre-date therapy, for example, childhood traumas (Gisquet-Verrier and Riccio, 2018;Houle and Philippe, 2020). Whenever possible, the interpretive context for memory integration should be positive self-representations, especially for negative experiences (Gisquet-Verrier and Riccio, 2018). For example, a memory of feeling lonely during childhood should be integrated through theory B ('unfairly treated'), rather than theory A ('useless and worthless'). For memories of positive experiences, the aim is to reflect, elaborate and consolidate them so that they become more memorable and accessible, when possible making explicit links to self-representations.

Risk reduction
The aim is to reduce suicide risk when it is heightened. Clients' motivation is explored in detail by asking about the intended and unintended consequences of suicidal actions. When feeling suicidal, clients' attention often narrows around a specific need, for example, to be re-united with a loved one, to escape, to experience relief or put an end to a particular emotion. The intention to die develops as a way of satisfying that need, although this is often implicit, confused and obscured by negative affect and rumination. Therapy makes the need explicit and invites reflection on whether suicide is the only or best way of satisfying that need. The need is taken seriously by considering non-lethal alternatives. In most cases, unintended consequences have not been fully considered and, when they are explored in detail, the client begins to reappraise whether suicide is a necessary or helpful path (e.g. pain, injury, illness, disability, distress, loss, guilt; Britton et al., 2011;Hawton et al., 2013). Suicide is usually motivated by some form of avoidance and clients are encouraged to switch to approach mode: to take their needs seriously and find other means of connection, escape, relief, etc. This process is augmented by broadening attention onto other life goals and reasons for living (Linehan et al., 1983). The final part of the process is making a safety plan: sharing information with others and preparing how to respond when risk increases. Clients need to access this learning when they feel suicidal, so that suicide is appraised as a bad strategy and non-lethal actions are chosen instead.

Staying well
This component is intended to counteract depressive relapse and it is usually the focus of booster sessions delivered towards the end of treatment (Jarrett et al., 2001). The client is encouraged to review and synthesise their learning across therapy, making explicit and consolidating their self-regulation skills. They are also encouraged to accept future challenges, setbacks and dysphoric moods as a normal part of life, rather than trying to avoid situations that could trigger disappointment or conflict (Jarrett and Thase, 2010). In preparation for treatment ending, it is helpful for clients to be exposed to mild stressors or setbacks, so they can learn how their mood is affected and practise self-regulation skills in an independent way. The aim is for knowledge and skills to be accessed when they are most needed, when clients start to become depressed. Ideally, clients learn that they can stay well without the continued presence of the therapist, and this limits the risk of relapse and future recurrences.

Relationship to other CBT therapies
The self-regulation model provides an integrated approach to treatment within the field of CBT. Alliance building, risk reduction and relapse prevention are similar to standard approaches, although they are delivered in a way that coheres with the self-regulation formulation. Alliance-building is emphasised more than in standard CT or BA, because of the high likelihood of alliance barriers in difficult-to-treat cases. Attending to mood fluctuations is also part of other therapies, for example, CT identifies changes in cognition when there are mood shifts. In self-regulation CBT (SR-CBT) it plays an even more central role, for two reasons: firstly, sometimes there are only micro-differences in mood across time, particularly in chronic cases, so greater attention needs to be paid to noticing fluctuations; secondly, less-depressed moods are essential for encouraging hope. Attention needs to be placed on less-depressed moods so that clients learn to associate them with motivation, engagement, reflection, positive memories and active life goals, even when less-depressed moods are fleeting at first. Like CT, mental freedom involves cognitive re-structuring and the emphasis is on establishing a better self-mind relationship through helpful processes such as attention, reflection and questioning. Unlike CT, negative thoughts and beliefs are not targeted directly, rather, an improved self-mind relationship makes it less likely that negatively biased thoughts will be generated. There are similarities with both mindfulness-based cognitive therapy (MBCT; Teasdale et al., 2000) and rumination-focused cognitive behavioural therapy (RFCBT; Watkins et al., 2011); however, in SR-CBT there is a distinctive emphasis on encouraging clients to mentally engage with issues that are subjectively important to them, learning how to reflect, question and make decisions about them effectively. Rumination is discouraged, thinking is not. There is also an emphasis on generating reasons to act with a lot of attention placed on clients' needs. There are points of contact with both motivational interviewing and some aspects of CFT, as attending to needs is a self-compassionate process and is much-needed in clients with self-critical rumination.
Like BA (and activity scheduling in CT), active engagement seeks to encourage behaviour change, but emphasises elucidating and satisfying the individual's needs and desires. Contextual factors are considered, but in service of needs and desires rather than in and of themselves. Like CT, change is targeted through behavioural experiments, but expectations and predictions (often negative) are not elicited. Instead, therapeutic time is devoted to eliciting preferred outcomes and exploring ways of influencing them. A desired outcome is a goal, not just a sequence of actions, and when preferred outcomes are influenced, the learning process is confirmatory rather than disconfirmatory. In this model, being able to influence or attain a goal is as important as having an enjoyable experience.
Active engagement and goal organisation have overlaps with values-based BA (Kanter et al., 2012), acceptance and commitment therapy (ACT; Zettle, 2007) and augmented depression therapy (ADepT; Dunn et al., 2019) in terms of committing to valued action and seeking to enhance positive affect, not just reduce negative affect. However, the emphasis in SR-CBT is on needs and desires rather than values. The difference is not just semantic: when people are depressed, values can be filtered through what they believe they should value, rather than what they actually want. For some clients, the question of what they want is new and their desires are an unchartered territory.
The need for memory integration is an acknowledgement of the prominence of intrusive memories in depression, and the high prevalence of both childhood and adult trauma in difficult-to-treat cases (Gisquet-Verrier and Riccio, 2018). In some cases, reliving is needed because of the intrusive and depressing effects of re-experiencing. This is not a novel therapy process but, to our knowledge, it is the first time it has been made explicit in a CBT treatment model for depression. Given the bespoke nature of treatment plans, memory integration can play a prominent role for some clients and a minor role for others.
Finally, self-organisation has overlaps with schema-focused work in CT, but the emphasis is different (Renner et al., 2018). The model does not target negative schemata but, when necessary, it will formulate how they developed to help a client de-centre from them. SR-CBT is less concerned with negatives and more concerned with positive self-representations: whether they are under-developed, disrupted or restricted in range, seeking ways to diversify and restructure self-identity so that vulnerability to depression is reduced.
Data availability statement. This article introduces a theoretical model: data availability is not applicable as no new data were created or analysed.